Available now

Qwen Lighting Restoration: lighting correction

By Alibaba

Qwen Lighting Restoration evens out the light in a photo, softening hard shadows and bright spots without a prompt.

Model ID
Maker
Alibaba

Input

Fields marked * are required. Review the price before generating.

The image to restore lighting for.

Uploading a file needs an account. Create account

JPEG, PNG, WebP · up to 25 MB

If not provided, the size of the final input image will be used.

Additional settings

The CFG (Classifier Free Guidance) scale. Controls how closely the model follows the prompt.

The number of inference steps to perform.

Acceleration level for image generation. 'regular' balances speed and quality.

Random seed for reproducibility. Same seed with same prompt will produce same result.

Price$0.06 / megapixel

You sign up first. Your settings are kept.

About this model

Overview

Upload one image and get it back with softer, more even lighting across the scene. There is no prompt to write; you can choose the output size, the number of images up to four, and the file format.

What you can build

  • Portraits taken in harsh midday sun, made softer
  • Product photos without hot spots from a flash
  • Even, neutral source images for further editing

Prompting tips

  • Use it on photos where the problem is the light, not the focus; blur and noise stay as they are.
  • Pick an output size close to the source shape so the framing does not change.

What to check

Check highlights, shadow transitions and color casts; generated restoration cannot prove the original lighting.

Inputs

  • Images (required): The URL of the image to restore lighting for.
  • Image Size (optional): The size of the generated image. If not provided, the size of the final input image will be used. Options: Square HD, Square, Portrait 4:3, Portrait 16:9, Landscape 4:3, Landscape 16:9.
  • Guidance Scale (optional): The CFG (Classifier Free Guidance) scale. Controls how closely the model follows the prompt. From 0 to 20.
  • Number of Inference Steps (optional): The number of inference steps to perform. From 2 to 50.
  • Acceleration (optional): Acceleration level for image generation. 'regular' balances speed and quality. Options: none, regular.
  • Negative Prompt (optional): The negative prompt for the generation.
  • Seed (optional): Random seed for reproducibility. Same seed with same prompt will produce same result.
  • Output Format (optional): The format of the output image. Options: png, jpeg, webp.
  • Number of Images (optional): Number of images to generate. From 1 to 4.

Output and limits

  • Image sizes: Square HD (1024×1024), Square (512×512), Portrait 4:3 (768×1024), Portrait 16:9 (576×1024), Landscape 4:3 (1024×768), Landscape 16:9 (1024×576).
  • File formats: PNG, JPEG, WEBP.
  • Up to 4 images per run.
  • Small or dense text in the image can come out misspelled.
  • Hands, crowds and complex poses can show artifacts; generate a few variations and pick the best.
  • Some prompts or outputs can be refused by safety checks.

Price guide

  • 1 megapixel: $0.06 / megapixel